skills/codervisor/forge/leanspec-sdd/Gen Agent Trust Hub

leanspec-sdd

Pass

Audited by Gen Agent Trust Hub on Mar 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the 'lean-spec' CLI for project management, including commands for viewing the project board, searching specs, and updating metadata. It includes a helper script 'scripts/validate-spec.sh' that executes the CLI via 'node' or 'npx'.
  • [COMMAND_EXECUTION]: The workflow documentation includes instructions for using 'git worktree' commands to handle parallel feature development in isolated directories.
  • [EXTERNAL_DOWNLOADS]: The skill references installation via the 'lean-spec' tool and 'npm'. As these resources (e.g., 'codervisor/forge') are provided by the skill's author or official registries, they are documented as standard configuration steps.
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection.
  • Ingestion points: The agent reads data from the 'specs/' directory via 'view' and 'search' commands.
  • Boundary markers: No specific delimiters or 'ignore' instructions are provided for interpolated specification content.
  • Capability inventory: The agent has access to 'lean-spec' CLI commands, 'git' commands, and local shell script execution.
  • Sanitization: No explicit sanitization of specification file content is defined within the skill.
  • Mitigation: The skill includes a 'Verify Against Reality' principle, directing the agent to validate status against the actual codebase and git history rather than relying solely on spec content.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 4, 2026, 07:16 AM